Exploring the impact of recreation on deer movement

Nikita Krähenbühl

Context: Deer movement data from Sihlwald

  • Application area: Recreation and Wildlife

  • Conceptual models:

    • Entity (movement of a single deer)

    • Field based (grouping of deer)

  • Data structures:

    • GPS Time Series (vector, unconstrained, active, regular measurements)
    • Hiking paths (vector)
    • Heatmap from strava (Raster)
  • Available data

    • 14 GPS Tracks over multiple years

    • Hiking trails, road infrastructure

    • Relative activity frequency data from Strava 1

Research questions

  • Does deer movement increase during periods of greater recreational activity
    • Weekend vs weekday

    • Evening vs midday vs early morning

  • Is deer movement constrained by recreational activity?
    • Do deer cross hiking paths, does this vary temporally?

Research plan

  • Perform explorative analysis

    • How many deer?
    • What timeframe?
    • Identify what to clean/filter
  • Gather additional data

Preliminary Results

Varied timeframes

How to continue?

  • Filters:
    • Timelag ~ 180 (deer overall)
    • Timelag ~ 3 minutes (deer short)
    • Steplength < 2500m
    • Remove inactive periods

Lets have a look at mean speed - averages per day, month and year

Mean speed averaged per day for 180 minute Intervals

Shorter intervals:

Spatial component: Hiking Paths?

Strava data: which paths are significant for sports?

Problems

  • I don’t think I will find weekend differences
  • Need to still explore constraints caused by hiking paths and examine Stava Data
    • Bit behind (hoping for a productive Hackathon)
  • Quarto and revealjs is tricky